海天背景下红外舰船目标实时跟踪算法研究

王悦行; 陈晓冬; 赵克军; 黄庆; 汪毅; 郁道银

doi:10.15892/j.cnki.djzdxb.2016.01.041

2016 , Vol. 36 >Issue 1: 165 - 170

DOI: https://doi.org/10.15892/j.cnki.djzdxb.2016.01.041

摘要

针对海天背景下的红外舰船图像对比度低、船载摄像机上下抖动剧烈的特点,文中提出了均值移位(mean-shift)和海天线检测相结合的红外目标实时跟踪算法。首先通过水平投影方法提取每帧图像的海天线,然后由相邻帧海天线的高度差设定每帧图像跟踪起始点,将目标的灰度和梯度信息进行加权来描述目标特征,利用mean-shift算法进行目标跟踪。实验结果表明,文中算法能够准确跟踪上下运动剧烈的红外舰船目标,跟踪速度达到12fps。

关键词： 均值移位; 红外舰船目标; 实时跟踪; 海天线; 水平投影; 加权直方图

本文引用格式

王悦行 , 陈晓冬 , 赵克军 , 黄庆 , 汪毅 , 郁道银 . 海天背景下红外舰船目标实时跟踪算法研究[J]. 弹箭与制导学报, 2016 , 36(1) : 165 -170 . DOI: 10.15892/j.cnki.djzdxb.2016.01.041

Abstract

A real-time algorithm under various complicated sea-sky background, which corresponded to the feature of low contrast around infrared ship and sharp up-and-down motion of camera on boat, was proposed by combining mean-shift and sea-sky-line detection.Firstly, the height of sea-sky-line was extracted by the horizontal projection method, and then the height difference of sea-sky-line in adjacent frames was used to set starting position at current frame, furthermore, the target's feature was described by creating weighted histogram which combined gray-level and gradients orientation information, and finally, mean-shift was used to track the position of target.Experimental results show that this algorithm can track infrared up-and-down motion ship target accurately, and the tracking speed can reach 12 fps.

Key words： mean-shift; infrared ship target; real-time tracking; sea-sky-line; horizontal projection; weighted histogram

[an error occurred while processing this directive]

参考文献

[1]	杨威, 李俊山, 史德琴. 时空联合的红外运动目标提取算法[J]. 光电工程, 2008, 35(5):50-54.
[2]	曾文静, 万磊, 张铁栋,等. 复杂海空背景下弱小目标的快速自动检测[J]. 光学精密工程, 2012, 20(2):403-412.
[3]	黄英东, 范宁军, 李杰. 一种基于海天线检测的舰船定位方法. 弹箭与制导学报, 2008, 28(5):286-288.
[4]	黄文韵, 马惠敏, 王生进. 海面背景红外目标的识别算法. 清华大学学报(自然科学版), 2009, 49(10):1609-1613.
[5]	KANG W J, DING X M, CUI J W, et al. Research on extraction of ship target in complex sea-sky background[C]// The International Symposium on Instrumentation Science and Technology, Journal of Physics:Conference Series, 2006, 48(1):354-358.
[6]	FEFILATYEV S, GOLDGOF D, SHREVE M, et al. Detection and tracking of ships in open sea with rapidly moving buoy-mounted camera system[J].Ocean Engineering, 2012, 54:1-42.
[7]	WU Jiawei, MAO Shiyi, WANG Xiaoping, et al. Ship target detection and tracking in cluttered infrared imagery[J].Optical Engineering, 2011, 50(5):(057207).
[8]	CHENG Yizong. Mean Shift, Mode Seeking, and Clustering[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(8):790-799.
[9]	朱胜利, 朱善安, 李旭超. 快速运动目标的 Mean shift 跟踪算法[J].光电工程, 2006, 33(5):66-70.
[10]	HUANG S, HONG J. Moving object tracking system based on camshift and kalman filter[C]//Proceedings of the International Conference on Consumer Electronics, Communications and Net-works.Piscataway:IEEE, 2011:1423-4426.
[11]	刘士建, 吴滢跃, 蔡能斌. 低 SNR 海天线提取算法[J].红外与激光工程, 2013, 42(12):3491-3495.
[12]	张旭光, 赵恩良, 王延杰. 基于 Mean-shift 的灰度目标跟踪新算法[J].光学技术, 2007, 33(2):226-229.
[13]	FOUAD Bousetouane, LYNDA Dib, HICHEM Snoussi. Improved mean shift integrating texture and color features for robust real time object tracking[J].The Visual Computer-Springer, 2013, 29(3):155-170.

Options

文章导航